Silencing of the MEKK2/MEKK3 Pathway Protects against Spinal Cord Injury via the Hedgehog Pathway and the JNK Pathway.

Spinal cord injury (SCI) is a devastating medical condition, often accompanied by motor and sensory dysfunction. The Hedgehog (Hh) pathway has a protective role in pathological injury after SCI. However, the specific mechanism remains unclear. The present study aimed to confirm the effects of the mitogen-activated protein kinase kinase-2 (MEKK2)/MEKK3/JNK/Hh pathway on SCI. SCI rat models were established and then inoculated with plasmids overexpressing MEKK2/MEKK3 or with small interfering RNA (siRNA) against MEKK2/MEKK3. The expression of MEKK2 and -3 was detected in dorsal root ganglia (DRG) cells. The motor function of hindlimbs, the expression of the c-Jun N-terminal kinase (JNK)- and Hh-pathway-related genes, and the level of neurofilament-200 (NF-200) and glial fibrillary acidic protein (GFAP) were measured. MEKK2 and -3 were expressed at a high level in DRG cells. The silencing of MEKK2/MEKK3 in rats caused an increase in the expression of glioma-associated oncogene homolog-1 (Gli-1), Nestin, smoothened (Smo), and Sonic Hedgehog (Shh). The Basso, Beattie, and Bresnahan (BBB) rating and the level of NF-200 protein also increased. However, the expression of monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein-1β (MIP-1β), MIP-3α, p-JNK/JNK, and p-c-Jun/c-Jun and the level of GFAP were reduced. Downregulation of MEKK2/MEKK3 ameliorated the symptoms of SCI by promoting neural progenitor cell differentiation via activating the Hh pathway and disrupting the JNK pathway. The findings in this study reveal a potential biomarker for SCI treatment.

Introduction

A spinal cord injury (SCI) can result in severe motor, sensory, and autonomic dysfunction. Each year, about 180,000 people worldwide sustain an SCI. The pathology of SCI is classified as a primary or secondary injury. A primary injury is defined as a mechanical event that directly affects the spinal cord, and a secondary injury is related to a complex cascade of molecular events, such as ischemia and inflammatory responses. In general, SCI is mainly caused by mechanical deformation of the adjacent tissues of the spinal column, i.e., the bony vertebrae, the spinal ligaments, and the intervertebral disc tissues. Until now, effective therapies for SCI have been limited, although antioxidant protection or engineered biomolecules have been used as treatments.5, 6 Interestingly, Sonic Hedgehog (Shh) ligand signaling increased the number of oligodendrocytes after SCI, induced proliferation of neural precursors, promoted formation of myelination, and assisted in nerve repair.

The Hedgehog (Hh) signaling pathway is recognized as critical in embryo patterning and development. In the development of the embryo, the Hh signaling pathway serves as both a mitogen and a morphogen that regulates the growth and patterning of multiple tissues, including the limbs and neural tube. The Hh pathway is recognized to consist of three Hh ligands in mammals: Sonic Hedgehog (Shh), Desert Hedgehog (Dhh), and Indian Hedgehog (Ihh). A previous study has revealed that the Shh signaling pathway regulates SCI by modulating the proliferation of neural and oligodendrocyte precursors. Zhang et al. have indicated that the activation of the Hh signaling pathway reduces inflammation in SCI and promotes recovery. MEKK3 a member of the mitogen-activated protein kinase (MAP3K) family, modulates various inflammatory and immune responses. The intracellular kinase MEKK2 forms a member of the MEKK/serine-threonine kinase 11 (STK11) family of MEKK and is an upstream regulator of c-Jun N-terminal kinase (JNK). Accumulating evidences suggests that the function of MEKK2 and -3 depends on the Hh signaling pathway. Both MEKK2 and -3 have been found to regulate JNK.15, 16 JNK belongs to a group of MAP kinases (MAPKs) and is upregulated when cells are subjected to environmental stresses. Previous research has shown that JNK has protective effects against pathological injury after SCI.18, 19 Until now, few studies have focused on the relationship of Hh and the MEKK2/MEKK3-JNK signaling pathway in SCI. Therefore, this study was conducted to identify the mechanism of SCI via the Hh and MEKK2/MEKK3-JNK signaling pathways.

Results

MEKK2 and MEKK3 Are Highly Expressed in DRG Cells

Western blot analysis was used to determine the expression of MEKK2 and -3 in lentivirus-infected dorsal root ganglion (DRG) cells. The results showed that the expression of MEKK2 and MEKK3 significantly increased in the MEKK2-, MEKK3-, and MEKK/MEKK3-vector groups (all p < 0.05), compared with that in the normal controls, whereas it decreased in the small interfering-MEKK2 (si-MEKK2), si-MEKK3, and si-MEKK2/MEKK3 groups (all p < 0.05; Figure 1).

Figure 1

Western Blot Analysis Revealed Increased Expression of MEKK2 and -3 in DRG Cells

(A) Western blots of MEKK2 and -3; (B) protein levels of MEKK2; and (C) protein levels of MEKK3. *p < 0.05 versus the normal controls; n = 3. The data are expressed as the mean ± SD, and compared by one-way ANOVA. The experiments were repeated three times.

Successful Establishment of the SCI Rat Model

In this experiment, the Basso, Beattie, and Bresnahan (BBB) locomotor rating scale was used to evaluate the motor function of the rats’ hindlimbs. As shown in Figure 2A, rats in the SCI group displayed a significant decrease in the BBB locomotor rating, accompanied with motor function loss to hindlimbs compared with that of the normal controls (p < 0.05), whereas no difference was observed in the sham injury group. In particular, rats with motor function loss could move only by using their forelimbs. However, the motor function of the hind limbs gradually recovered, resulting in an increase in the BBB locomotor rating.

Figure 2

BBB Score and H&E Staining Confirmed the Successful Establishment of an SCI Rat Model

(A) Motor function of rat hindlimbs at days 0, 7, 14, 21, and 28 after surgery evaluated according to the BBB scale. (B) Histopathological changes in spinal cords of rats at days 0, 7, 14, 21, and 28 after surgery detected by H&E staining (×400). (C and D) The positive levels of GFAP (C) and NF-200 (D) and in spinal cord tissues of rats at day 28 after surgery, evaluated by immunohistochemistry (×400). *p < 0.05 versus the normal group. The data are expressed as the mean ± SD. Data in (A) were compared by repeated-measures ANOVA and in (B)–(D) by unpaired t test; n = 10.

After establishment of the SCI rat model, H&E staining was performed to detect the histopathological changes in the spinal cord. As shown in Figure 2B, spinal cord tissues in the SCI group displayed a large number of cavities with numerous necrotic tissue fragments. Liquefactive necrosis and edematous white matter (sponge-like) were also observed. However, neurons and glial cells in the normal controls exhibited a uniform distribution with a plump cell body and a clear boundary between nucleus and cytoplasm.

The neurofilament protein NF-200 is present in the axon under normal conditions, but not in the cell body. When the spinal cord is injured, the adjacent neurons not only reflect the functional status of the neurons as a result of trauma and various inducing factors, but also indicate the condition of axon regeneration. The glial fibrillary acidic protein (GFAP) is specific to the nervous system and is expressed only in the soma, dendrites, and axons of the spinal cord and brain astrocytes. It can be used to detect the differentiation of neural stem cells in the spinal cord after injury.20, 21 After establishment of the SCI rat model, immunohistochemistry was performed to detect the levels of NF-200 and GFAP in the spinal cord. As depicted in Figures 2C and 2D, the levels of NF-200 and GFAP in the sham injury group did not differ significantly from those in the normal control group (p > 0.05), whereas the SCI group showed a reduced level of NF-200 and an increased level of GFAP (p < 0.05).

Inhibition of the Hh Pathway Inhibits Recovery from SCI

Compared with the sham injury group, the SCI group showed an increase in mRNA expression in glioma-associated oncogene homolog-1 (Gli-1), Nestin, smoothened (Smo), and Shh (all p < 0.05). However, the SCI+cyclopamine group exhibited a significant decrease in expression of Gli-1, Nestin, Smo, and Shh mRNA, compared with the SCI+DMSO group (all p < 0.05; Figures 3A and 3C). These results were also confirmed by western blot analysis (Figures 3B and 3D). Meanwhile, the BBB scale was applied to evaluate the motor function of hindlimbs in the sham, SCI, SCI+DMSO, and SCI+cyclopamine groups. As shown in Figure 3E, rats in the SCI+cyclopamine group showed an obvious decrease in BBB locomotor ratings compared with those in the SCI+DMSO group (p < 0.05). Results from the H&E staining showed that the spinal cord tissue in the SCI+cyclopamine group exhibited a larger number of cavities with necrotic tissue fragments and liquefactive necrosis than in the SCI+DMSO group (Figure 3F). Meanwhile, immunohistochemistry results showed that the level of NF-200 decreased and the level of GFAP increased in the SCI+cyclopamine group compared with levels in the SCI+DMSO group (all p < 0.05; Figure 3G).

Figure 3

Suppression of the Hh Pathway Prevents SCI from Recovery

(A) mRNA expression of Gli-1, Nestin, Smo, and Shh in the sham injury and SCI groups at day 28 after surgery, determined by qRT-PCR. (B) Protein levels of Gli-1, Nestin, Smo, and Shh in the sham injury and SCI groups at day 28 after surgery, determined by western blot analysis. (C) mRNA expression of Gli-1, Nestin, Smo, and Shh in each group at day 28 after surgery, determined by qRT-PCR. (D) Protein levels of Gli-1, Nestin, Smo, and Shh in each group at day 28 after surgery, determined by western blot analysis. (E) Motor function of rats’ hindlimbs in each group at days 0, 7, 14, 21, and 28 after transduction, evaluated by BBB scale. (F) Histopathological changes of spinal cord tissues in each group at day 0, 7, 14, 21, and 28 after transduction was detected by H&E staining (×100). (G) Positive levels of NF-200 and GFAP in each group at day 28 after transduction, determined by immunohistochemistry (×400). *p < 0.05 versus the sham injury/SCI+DMSO group. The data are expressed as the mean ± SD. Data in (E) were compared by repeated-measures ANOVA and in the other panels by unpaired t test; n = 10.

Inhibition of the JNK Pathway Has Protective Effects on SCI

qRT-PCR revealed that the SCI group displayed a notable elevation regarding mRNA expression of monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein-1β (MIP-1β), and MIP-3α, compared with the sham injury group (all p < 0.05). However, a decrease was observed in the SCI+SP600125 group in comparison with the SCI+DMSO group (all p < 0.05; Figures 4A and 4C). Meanwhile, western blot analysis revealed that the expression of p-JNK/JNK and p-c-Jun/c-Jun increased significantly in the SCI group compared with that in the sham injury group (all p < 0.05; Figure 4B). An opposite trend was observed in the SCI+SP600125 group in comparison with the SCI+DMSO group (all p < 0.05; Figure 4D).

Figure 4

Disruption of the JNK Pathway Has Protective Effects against SCI

(A) mRNA expressions of MCP-1, MIP-1β, and MIP-3α in the sham injury and SCI groups at day 28 after surgery, determined by qRT-PCR. (B) Protein levels of p-JNK/JNK and p-c-Jun/c-Jun in the sham injury and SCI groups at day 28 after surgery, detected by western blot analysis. (C) mRNA expression of MCP-1, MIP-1β, and MIP-3α in each group at day 28 after surgery detected by qRT-PCR. (D) Protein levels of p-JNK/JNK and p-c-Jun/c-Jun in each group at day 28 after surgery, detected by western blot analysis. (E) Motor function of rat hindlimbs in each group at days 0, 7, 14, 21, and 28 after transduction evaluated by BBB rating. (F) Histopathological changes of spinal cord tissues in each group at day 28 after transduction determined by H&E staining (×100). (G) The positive levels of NF-200 and GFAP proteins in each group at day 28 after transduction, determined by immunohistochemistry (×400). *p < 0.05 versus the sham/SCI+DMSO group. The data are expressed as the mean ± SD. Data in (E) were compared by repeated-measures ANOVA and in the other panels by unpaired t test; n = 10.

In addition, the BBB scale was used to determine the motor function of hindlimbs of rats in each group. The results showed that rats in the SCI+SP600125 group had obviously increased ratings on the BBB locomotor scale in contrast to those of the SCI+DMSO group (p < 0.05; Figure 4E).This result was also confirmed by H&E staining which revealed that the spinal cord tissues of rats in the SCI+SP600125 group exhibited a decreased number of cavities and less necrosis compared with those in the SCI+DMSO group (Figure 4F). Meanwhile, immunohistochemistry results showed that the expression level of NF-200 increased and that of GFAP decreased in the SCI+SP600125 group, compared with the SCI+DMSO group (all p < 0.05; Figure 4G). All the above results suggest that inhibition of the JNK pathway may confer protective effects on SCI.

Overexpression of MEKK2/MEKK3 Aggravates SCI by Inhibiting the Hh Pathway and Activating the JNK Pathway

The SCI rats received a spinal cord injection of MEKK2 and -3 vectors, in order to investigate the relations of MEKK2/MEKK3 to the Hh and JNK pathways in SCI. Western blot analysis showed that a distinctive increment of MEKK2/MEKK3 expression was found in the tissues of injured spinal cords (Figure 5A). qRT-PCR results showed that expression of Gli-1, Nestin, Smo, and Shh mRNA decreased significantly in the SCI+MEKK2-, SCI+MEKK3-, and SCI+MEKK/MEKK3-vector groups compared with the SCI+NC-vector group (all p < 0.05), whereas the expression of MCP-1, MIP-1β, and MIP-3α increased (all p < 0.05; Figure 5B). Similarly, western blot analysis revealed that expression of Gli-1, Nestin, Smo, and Shh decreased significantly in the SCI+MEKK2-, SCI+MEKK3-, and SCI+MEKK/MEKK3-vector groups compared with that of the SCI+NC-vector group (all p < 0.05), whereas p-JNK/JNK and p-c-Jun/c-Jun increased (all p < 0.05; Figures 5C and 5D).

Figure 5

Overexpressed MEKK2/MEKK3 Aggravates SCI by Inhibiting the Hh Pathway and Activating the JNK Pathway

(A) The transduction efficiency of MEKK2/MEKK3 vectors, determined by western blot analysis. (B) mRNA expression of Hh- and JNK-pathway-related factors in each group at day 28 after surgery, detected using qRT-PCR. (C) Protein levels of Hh-pathway-related factors in each group at day 28 after surgery, measured by western blot analysis. (D) Protein levels of JNK pathway-related factors in each group at day 28 after surgery, detected using western blot analysis. (E) Motor function of the hindlimbs of rats in each group at days 0, 7, 14, 21, and 28 after transduction, tested according to the BBB scale. (F) Histopathologic changes in spinal cord tissues in each group at day 28 after transduction, observed with H&E staining (×100). (G) Positive level of NF-200 and GFAP proteins in each group at day 28 after transduction, determined by immunohistochemistry (×400). *p < 0.05 versus the SCI+NC-vector group. The data are expressed as the mean ± SD. Data in (E) were compared by repeated-measures ANOVA and in the other panels by one-way ANOVA; n = 10.

Meanwhile, rats in the SCI+MEKK2, SCI+MEKK3, and SCI+MEKK/MEKK3 vector groups showed an obvious decrease in BBB locomotor rating scale compared with that of the SCI+NC-vector group (p < 0.05; Figure 5E). H&E staining revealed that rats in the SCI+MEKK2-, SCI+MEKK3-, and SCI+MEKK/MEKK3-vector groups exhibited an increased number of cavities and more liquefactive necrosis in comparison with the SCI+NC-vector group (Figure 5F). Immunohistochemistry results showed that the level of NF-200 was reduced and that of GFAP was elevated in the SCI+MEKK2-, SCI+MEKK3-, and SCI+MEKK/MEKK3-vector groups compared with the SCI+DMSO group (all p < 0.05; Figure 5G). These results demonstrate that overexpressed MEKK2/MEKK3 may exacerbate SCI by inhibiting the Hh pathway and activating the JNK pathway.

Silencing of MEKK2/MEKK3 Abates SCI by Activating the Hh Pathway and Inhibiting the JNK Pathway

MEKK2/MEKK3 expression in spinal cord tissues of SCI rats receiving si-MEKK2/MEKK3 was detected by means of western blot analysis, aimed at analyzing the silencing efficiency. A notable decline in MEKK2/MEKK3 expression was noted in the spinal cord tissues of SCI rats after injection (Figure 6A). qRT-PCR and western blot analysis results confirmed that mRNA and expression of the Gli-1, Nestin, Smo, and Shh proteins in the SCI+si-MEKK2, SCI+si-MEKK3, and SCI+si-MEKK2/MEKK3 groups increased significantly compared with that of the SCI+si-NC group (all p < 0.05), whereas the expression of MCP-1, MIP-1β, and MIP-3α was reduced (all p < 0.05). qRT-PCR also showed that, compared with the SCI+si-MEKK2/MEKK3+DMSO group, in the SCI+si-MEKK2/MEKK3+cyclopamine group the expression of Gli-1, Nestin, Smo, and Shh mRNA showed a decrement (all p < 0.05), whereas that of MCP-1, MIP-1β, and MIP-3α remained statistically similar (all p > 0.05). Western blot analysis also showed that, compared with the SCI+si-MEKK2/MEKK3+DMSO group, the SCI+si-MEKK2/MEKK3+cyclopamine group had significantly decreased expression of Gli-1, Nestin, Smo, and Shh (all p < 0.05), whereas the expression of p-JNK/JNK and p-c-Jun/c-Jun showed no difference (all p > 0.05; Figures 6B–6D).

Figure 6

MEKK2/MEKK3 Disruption Alleviates SCI by Activating the Hh Pathway and Inhibiting the JNK Pathway

(A) The transduction efficiency of si-MEKK2/MEKK3, determined by western blot analysis. (B) mRNA expression of Hh- and JNK pathway-related factors in each group at day 28 after transduction, evaluated by qRT-PCR. (C) Protein levels of Hh pathway-related factors in each group at day 28 after transduction, measured by western blot analysis. (D) Protein levels of JNK pathway-related factors in each group at day 28 after transduction, measured by western blot analysis. (E) motor function of the hindlimbs of rats in each group at days 0, 7, 14, 21, and 28 after transduction, evaluated according to the BBB scale. (F) Histopathological changes in spinal cord of rats in each group at day 28 after transduction, according to H&E staining (×400). (G) Positive level of NF-200 and GFAP proteins in each group at day 28 after transduction, determined by immunohistochemistry (×400). *p < 0.05 versus the SCI+si-NC group; #p < 0.05 versus the SCI+si-MEKK2/MEKK3+DMSO group. The data are expressed as the mean ± SD. Data in (E) were compared by repeated-measures ANOVA and in the other panels by one-way ANOVA; n = 10.

The BBB scale was used to test the motor function of the hindlimbs of rats. As shown in Figure 6E, compared with the SCI+si-NC group, rats in the SCI+si-MEKK2, SCI+si-MEKK3, and SCI+si-MEKK2/MEKK3 groups had a markedly elevated BBB locomotor rating (p < 0.05). However, the BBB locomotor rating decreased in the SCI+si-MEKK2/MEKK3+cyclopamine group in comparison with the SCI+si-MEKK2/MEKK3+DMSO group (p < 0.05).

H&E staining was performed to observe the histopathological changes in spinal cord tissues of rats in each group. As shown in Figure 6F, compared with the SCI+si-NC group, the SCI+si-MEKK2, SCI+si-MEKK3, and SCI+si-MEKK2/MEKK3 groups exhibited a reduced number of cavities and less liquefactive necrosis in the tissue sections. Nevertheless, an increased number of cavities in the tissue sections and more liquefactive necrosis in the tissues were observed in the SCI+si-MEKK2/MEKK3+cyclopamine group in contrast to the SCI+si-MEKK2/MEKK3+DMSO group (all p < 0.05).

Immunohistochemistry was used to detect a positive level of NF-200 and GFAP proteins. As shown in Figure 6G, compared with the SCI+si-NC group, the level of NF-200 protein markedly increased and that of GFAP decreased in the SCI+si-MEKK2, SCI+si-MEKK3, and SCI+si-MEKK2/MEKK3 groups (all p < 0.05), whereas the change was reciprocal in the SCI+si-MEKK2/MEKK3+cyclopamine group, compared with that of the SCI+si-MEKK2/MEKK3+DMSO group. Taken together, the evidence showed that silencing MEKK2/MEKK3 ameliorates SCI by activating the Hh pathway and inhibiting the JNK pathway.

Discussion

SCI is recognized as one of the most common reasons for paralysis around the world. It is accompanied by permanent neurological defects that derive from loss of neurons and axons. Therefore, SCI seriously affects the quality of life of patients and their families. The aim of the present study was to determine the modulatory effects of the MEKK2/MEKK3/JNK/Hh pathway in SCI.

Our data demonstrated that the Hh signaling pathway was upregulated in SCI rats. The Hh signaling pathway controls several processes that are deregulated in patients with metabolic syndrome, including vascular remodeling, carcinogenesis, liver injury, and repair. The Hh pathway consists of three Hh ligands in mammals: Shh, Dhh, and Ihh. Shh, a significant protein that participates in craniofacial morphogenesis, is secreted from reactive astrocytes after a cerebral cortex injury.23, 24 Shh releases the associated signaling receptor Smo, resulting in upregulation and nuclear translocation of Gli transcription factors. The Gli transcription factors are regarded as key mediators of activation of the Hh pathway, and their expression is upregulated in several human cancers, independent of the upstream Hh signaling pathway.

This study also confirmed that SCI rats exhibited overexpression of MCP-1, MIP-1β, MIP-3α, p-JNK/JNK, and p-c-Jun/c-Jun, suggesting that the JNK signaling pathway was upregulated in SCI rats. MCP-1 belongs to the β-chemokine family, which activates and recruits mononuclear phagocytes and is activated in SCI. MCP-1, expressed in the DRG, is upregulated in spinal cord astrocytes after nerve injury, and its upregulation is associated with JNK. In addition, MIP-1β and -3α are considered chemokines that have been found to participate in various inflammatory responses.29, 30 It has been proved that the JNK signaling pathway is involved in neurological inflammatory conditions. MIP-1α is produced by immune cells, including macrophages, lymphocytes, and dendrites. It can activate immunocytes, promote chemotaxis, and regulate the synthesis of cytokines and is involved in acute or mild inflammation.32, 33 Lee et al. proposed that JNK activation after SCI is involved in apoptotic neuronal cell death and axonal degeneration, leading to limited motor recovery after SCI. Our data are consistent with the above research, which showed that suppressed JNK could have a protective effect on SCI.

Cyclopamine acts as an inhibitor of the Hh pathway, which specifically binds and represses the Smo receptor. In our study, the cyclopamine group exhibited a significant decrease in Gli-1, Nestin, Smo, and Shh expression in SCI rats. Meanwhile, lower scores on the BBB scale, a decreased level of NF-200, and an increased level of GFAP were also observed. These results suggest that cyclopamine suppresses the Hh signaling pathway and exacerbates SCI. The Hh pathway is activated in nerve damage, and may serve as a potential biomarker for treating neurodegenerative diseases and dysfunction, including SCI. SP600125 is known as one of the JNK inhibitors. Our data showed that the SP600125 group exhibited significant decreases in MCP-1, MIP-1β, MIP-3α, p-JNK/JNK, and p-c-Jun/c-Jun expression, whereas higher ratings on the BBB scale, an increased level of NF-200, and a reduced level of GFAP indicated that SP600125 may protect against the effects of SCI by suppressing the JNK signaling pathway. JNK is one of the MAPK pathways that participates in SCI. A previous study has demonstrated that JNK suppression serves as an effective strategy for prevention of the evolution of secondary damage after SCI. The BBB scale was often used to evaluate injury severity in SCI rat models. Wang et al. observed that NF-200 started to decrease at 2 h and dropped to its lowest at 24 h. NF-200 expression rebounded from day 3 and reached original levels at day 28. The number of cells expressing GFAP increased from day 3, peaked at day 14, and began recovering at day 28. The changes of neuron-specific enolase (NSE), NF-200, GFAP, and cortical somatosensory evoked potentials (CSEPs) were significantly associated with the BBB score, suggesting that NF-200 and GFAP expression correlates with the BBB locomotor rating in response to SCI.

Moreover, we showed that silencing of MEKK2/MEKK3 could ameliorate SCI through activating the Hh pathway and inhibiting the JNK pathway. Rats in the MEKK/MEKK3-vector group had increased expression of Gli-1, Nestin, Smo, and Shh, whereas it decreased in the si-MEKK2/MEKK3 group. Lu et al. indicated that MEKK2/MEKK3 correlates negatively with the Hh signaling pathway in medulloblastoma. Moreover, deletion of MEKK2/MEKK3 leads to accumulation of Gli-1. It also has been reported that upregulated MEKK2 leads to JNK activation, suggesting that MEKK2 is positively related to JNK. As previously mentioned, activation of Hh and inhibition of JNK could ameliorate SCI. Therefore, we supposed that silencing MEKK2/MEKK3 could alleviate SCI via activating Hh and suppressing JNK.

We concluded that downregulation of MEKK2/MEKK3 could benefit recovery from SCI. This protective effect depends on the suppression of the Hh signaling pathway and activation of the JNK signaling pathway. Therefore, the MEKK2/MEKK3/JNK/Hh pathway may serve as a prognostic indicator for SCI treatment in the future. Further studies will be conducted to fully interpret specific mechanisms of the MEKK2/MEKK3-JNK/Hh signaling pathway in SCI treatment.

Materials and Methods

Ethics Statement

All animal experiments were performed in accordance with experimental animal surgical procedures, approved by the Ethics Committee of Shanghai East Hospital, School of Medicine, Tongji University.

Rat Model

A total of 180 Sprague-Dawley (SD) male rats weighing 150–200 g were used in this study and were classified into three groups: a normal control group (n = 10), a sham injury group (n = 10), and an SCI model group (n = 140). After being anesthetized with 3% pentobarbital sodium (P3761; Sigma-Aldrich, St. Louis, MO, USA), the rats were fixed in a prone position on the operating table, treated with surgical disinfectant, and covered with surgical drapes. The skin, subcutaneous tissues, and fascia of the spinal cord area (T7–T9 segment) were incised with a midline incision (3–4 cm), and the lateral muscle tissues were separated to expose the neural plate. Then, the spinous process and neural plate were disconnected from the spinal cord with micro-rongeur forceps. After it was exposed, the spinal cord was set at the top of a plastic rod (diameter, 3 mm) and destroyed by a vertically released stainless steel rod (diameter, 2.5 mm; weight, 5 g), which was set 5 cm away from the plastic rod. The spinal cord was then weighed and sutured with a no. 0 surgical needle. A successful SCI model was defined as a spinal cord with hemorrhage and edema, followed by contractions of the lower extremity, acutely swung tail, and flaccid paralysis. One hundred and forty SCI rat models were successfully established with an overall success rate of 87.50%. Ten rats with successfully induced SCI served as the SCI group in the subsequent experiments. After the establishment of the SCI model rat group, a gelatin sponge was used for hemostasis until no obvious bleeding was observed at the surgical site, followed by suturation of the incision. No spinal cord injury was imposed in the sham injury group, in which rats underwent only neural plate destruction and spinal cord exposure. All the surgical procedures were performed under aseptic conditions, and body temperature was maintained at 37°C with an electric heating plate. Surgically injured rats were protected against infection during the 1–3 days after surgery, and the bladder was gently pressed two to three times until spontaneous urination occurred.

Construction of Lentivirus

The overexpressed and silenced sequences of MEKK2 (ID26405) and MEKK3 (ID26406) were synthetized by Shanghai Sangon Biological Engineering Technology & Services (Shanghai, China). The overexpressed and silenced segments of MEKK2 and MEKK3 and the target vector pLVX-IRES-ZsGreenl (pLVX-shRNA) were treated with EcoRI and BamHI for double-enzyme digestion. The digested vectors and MEKK2 and MEKK3 segments were then connected and transformed into Escherichia coli DH5-α cells. Subsequently, the plasmid vectors with pspax2 (5 μg) and pMD2G (5 μg; Invitrogen, Carlsbad, CA, USA) were combined with recombinant plasmids pLVX-MEKK2-IRES-ZsGreenl (10 μg), pLVX-MEKK3-IRES-ZsGreenl (10 μg), pLVX-shRNA-MEKK2 (10 μg), or pLVX-shRNA-MEKK3 (10 μg). The hybrid was then extracted with a high-purity non-endotoxin, and co-transduced with 293T cells. Then, 8 h after transduction, the solution was replaced with complete medium. Meanwhile, the empty plasmids were co-transduced into 293T cells. Next, after an incubation period of 48–72 h, the supernatant containing recombinant lentivirus particles were collected and the concentration raised to a high titer. Meanwhile, the rat DRG cells were incubated with the above lentivirus and lysed after transduction for 48 h. The expression of MEKK2 and MEKK3 was then detected by immunoblot analysis, and the interference capacity of the lentivirus was defined. The remaining lentivirus with a concentration higher than 107 transduction units (TU)/mL was stored at −80°C for future use.

Isolation of DRG Cells

The DRG ganglion pieces were treated with 0.25% trypsin and then detached in a 37°C water bath for 20 min. Next, Neurobasal medium containing 10% fetal bovine serum was added to terminate the detachment. The DRG pieces were centrifuged at 1,000 rpm/min for 5 min and then resuspended in serum-free Neurobasal medium. After centrifugation at 1,000 rpm/min for 5 min, the cells were resuspended in Neurobasal medium, and the cell suspension was inoculated into a 24-well plate with poly-L-lysine-coated coverslips and incubated with 5% CO2 at 37°C for 4 h. Neurobasal medium (1 mL) was added to each well and then was replaced with 2% B-27 Neurobasal medium after 24 h, after which the culture medium was changed every 3 days.

Grouping and Transduction

The rest of the 130 SCI rats were classified into 13 groups (10 rats/group) and received a spinal cord injection according to group: SCI+DMSO group (spinal cord injection with DMSO, 5 μg per rat); SCI+SP600125 group (SP600125, 5 μg per rat); SCI+cyclopamine group (cyclopamine, 5 μg per rat); SCI+negative control (NC)-vector group (8 μL lentivirus containing empty plasmid vectors); SCI+MEKK2-vector group (8 μL lentivirus containing MEKK2 overexpressed plasmid vectors); SCI+MEKK3-vector group (8 μL lentivirus containing MEKK3 overexpressed plasmid vectors); SCI+MEKK2/MEKK3-vector group (8 μL lentivirus containing MEKK2/MEKK3 overexpressed plasmid vectors); SCI+si-NC group (8 μL lentivirus containing empty plasmid vectors); SCI+si-MEKK2 group (8 μL lentivirus containing MEKK2 silencing plasmid vectors); SCI+si-MEKK3 group (8 μL lentivirus containing MEKK3 silencing plasmid vectors); SCI+si-MEKK2/MEKK3 group (8 μL lentivirus containing MEKK2/MEKK3 silencing plasmid vectors); SCI+si-MEKK2/MEKK3+DMSO group (8 μL lentivirus containing MEKK2/MEKK3 silencing plasmid vectors and DMSO; 5 μg per rat); and SCI+si-MEKK2/MEKK3+cyclopamine group (8 μL lentivirus containing MEKK2/MEKK3 silencing plasmid vectors and cyclopamine, 5 μg per rat). All rats were fed with regular food with free access to water. Light and dark in the feeding room were alternated every 12 h.

Measurement of Rat Behavior

A behavioral test was performed on days 0, 7, 14, 21, and 28 after surgery to evaluate the motor function of the hindlimbs and the results were rated according to the BBB locomotor scale. In detail, rats were first placed in an open space (125 × 125 cm) for orientation. Then, motor function was observed and scored after 5 min by two persons who were familiar with the BBB scale and were not involved in this study. The BBB score was defined as the mean value of three recorded values, with 0 points representing complete paralysis and 21 points representing normal movement.

H&E Staining

At day 28 after surgery, spinal cord tissues were fixed in formaldehyde for 1–2 min and sliced into sections. After staining with hematoxylin for 3 min, sections were treated with 1% hydrochloric acid alcohol for 5 s and washed in running water for 5 min. After they were washed, the sections were stained with eosin for 1 min and dehydrated in gradient alcohols (70%, 75%, 80%, 90%, and 100%) for 2 min each. After treatment with dimethylbenzene for transparency, the sections were sealed with neutral balsam and observed under a light microscope. Five randomly selected fields were measured for each sample.

RNA Extraction and Quantification

Total RNA was extracted from the spinal cord tissues using the Ultrapure RNA Extraction Kit (QIAGEN, Hilden, Germany). The primers were designed and synthetized by AOKE Biotechnology (Guangzhou, China) (Table 1). qRT-PCR was performed with a SYBR Premix Ex Taq II kit (RR820A; Xingzhi Biotechnology, Guangzhou, China) and ABI Prism 7300 system (Prism 7300; Shanghai Kunke Instrument and Equipment, Shanghai, China). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH; abs830032, Absin Bioscience, Shanghai, China) was used as the internal reference. The ratio of relative gene expression between the SCI and normal groups was analyzed by the 2−ΔΔCt method.

Table 1

Primer Sequences of Related Genes for qRT-PCR

Gene	Primer Sequence
Shh	F: 5′-AGATGCTGCTGCTGCTGG-3′
	R: 5′-TCCTTCCGTGCGCTTTCC-3′
MCP-1	F: 5′-TCA GCC AGA TGC AGT TAA CG-3′
	R: 5′-GATCCTCTTGTAGCTCTC CAG C-3′
MIP-1β	F: 5′-TCCCAC TTC CTG CTG TTT CTCT-3′
	R: 5′-GAATAC CAC AGC TGG CTT GGA-3′
MIP-3α	F: 5′-GACTGCTGCCTCACGTACAC-3′
	R: 5′-CGACTTCAGGTGAAAGATGATAG-3
Gli-1	F: 5′-CTCTGCTGACTCTGGGATATG-3′
	R: 5′-GATCAGGATAGGAGCCTGCTG-3′
Nestin	F: 5′- GCGGGGCGGTGCGTGACTAC-3′
	R: 5′-AGGCAAGGGGGAAGAGAAGGATGT-3′
Smo	F: 5′- GCAGTTCCTCGGCTGCCTC −3′
	R: 5′-AGCCTCCATTAGGTTAGTGCG-3′
GRP94	F: 5′-GACGGGCAAGGACATCTCTACAAA-3′
	R: 5′-CTTCTTCTTCTGCCCCTGCGTCTG-3′
GAPDH	F: 5′-GCAAGTTCAACGGCACAG-3′
	R: 5′-AGGTGGAAGAATGGGAGTTGC-3′

Shh, Sonic Hedgehog ligand; MCP, monocyte chemoattractant protein; MIP, macrophage inflammatory protein; Gli-1, glioma-associated oncogene homolog-1; Smo, smoothened; GRP94, glucose-regulated protein 94; F, forward; R, reverse; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Western Blot Analysis

Spinal cord tissues were mixed in lysis buffer to collect total protein, and the protein concentration was determined by bicinchoninic acid (BCA) kit (20201ES76; YEASEN Biotechnology, Shanghai, China). The samples were then mixed with 10% SDS gel loading buffer and boiled at 100°C for 10 min. After centrifugation, the supernatant containing the proteins was loaded into each lane, separated by SDS-PAGE, and wet transferred to a nitrocellulose (NC) membrane, followed by blocking with 5% skimmed milk. After blocking, the membrane was incubated with primary antibodies (all purchased from Abcam, Cambridge, MA, USA) at 4°C overnight, including Shh (1:1,000, ab19897), Gli-1 (1:100,000, ab49314), GADPH (1:10,000, ab8245), Smo (1:1,000, ab5694), Nestin (1:1,000, ab6142,), p-JNK (1:1,000, ab124956), JNK (1:1,000, ab179461), p-c-Jun (1:1,000, ab32385), c-Jun (1:1,000, ab40766), MEKK2 (1:10,000, ab33918), and MEKK3 (1:5,000, ab26321). After it was washed with Tris-buffered saline-Tween-20 (TBST) three times for 5 min each, the membrane was incubated with horseradish-peroxidase (HRP)-labeled goat anti-rabbit or rat IgG (1:1,000; Boster Biological Technology, Wuhan, Hubei, China) for l h at 37°C. After three washes with TBST for 5 min each, the membrane was immersed in chemiluminescent substrate (ECL; Pierce Biotechnology, Waltham, MA, USA) solution for 1 min and subjected to chemiluminescence analysis (Shanghai Jing Ke Chemical Technology, Shanghai, China). GAPDH was used as the internal reference. The relative protein expression was defined as the ratio of the gray value of the target band to that of the internal reference.

Immunohistochemistry

The SP Immunohistochemistry Kit (HSP0001; Shanghai Maibio, Shanghai, China) was used in this experiment. The spinal cord tissue was fixed in formaldehyde, embedded in paraffin, and cut into 4 μm sections. Then, the sections were heated in a microwave for 1 h at 70°C, dewaxed in dimethylbenzene, and dehydrated in gradient alcohols. After incubation with 3% H2O2 for 10 min at room temperature, the sections were treated with goat serum (Beijing ComWin Biotech, Beijing, China). After blocking, three randomly selected sections in each group were incubated with the mouse monoclonal anti-NF-200 antibody (1:100) and rabbit polyclonal anti-GFAP antibody (AmyJet Scientific, Wuhan, China) at 4°C overnight. The sections were incubated with biotinylated rabbit anti-goat/goat anti-rabbit IgG (Boster Biological Technology, Wuhan, Hubei, China) for 20 min at 37°C, incubated with peroxidase-labeled streptavidin for 30 min at 37°C, visualized with diaminobenzidine (DAB; DA1010, 3 mL; Beijing Solarbio Science & Technology, Beijing, China) for 5–10 min, and finally observed under a light microscope. Five randomly selected fields were measured for each section. The criteria for scoring were on the basis of positive astrocytes and neurofilaments. When 50 high-magnification fields were counted, the percentage of positive cells per 1,000 cells was determined to indicate the level, where <50% was regarded as (−) and ≥50% as (+).

Statistical Analysis

All data are expressed as the mean ± SD. Statistical analysis was performed with SPSS 21.0 software (IBM, Armonk, NY, USA). The data with normal distribution and with homogeneity of variance were tested by unpaired t test. Comparisons among multiple groups were analyzed using one-way ANOVA, and the normality of the data was tested by using the Kolmogorov-Smirnov method. Comparisons of data with normal distribution were made by ANOVA, followed by Tukey’s multiple comparison post hoc tests. Moreover, the BBB scale at different time points was analyzed by repeated-measures ANOVA. p < 0.05 was statistically significant.

Author Contributions

Y.-D.Z., Y.-L.K., Y.-F.W. and Z.-S.Z. designed the study. Z.-W.D., D.Z., and J.C. collated the data and carried out data analyses. Q.-H.J. and S.-C.Z. produced the initial draft of the manuscript. Q.-H.J. and S.-C.Z. contributed to drafting the manuscript. All authors have read and approved the final submitted version.

Conflicts of Interest

The authors declare no competing interests.